Research On The Method For Swarm-robot Forming Formations

With the increase of task complexity and environmental uncertainty and variability,the limitations of single robot are becoming more and more prominent due to its own performance and reliability.Swarm-robot system which has advantages of robustness,self-repair ability and self-adaptability has attracted wide attention.As the foundation of other coordination problems,swarm-robot formation control is enormously worth to study.According to the problem-solving process,swarm-robot formation can be divided into formation forming and formation keeping.Formation keeping can be regarded as forming a formation in real-time,so the first problem of swarm-robot formation is forming a formation.Formation forming consists of two important parts.How should a robot choose its corresponding point when facing several target points.How to avoid confrontations with other robots within the process of forming formations.This paper solves swarm-robot formation from the above two problems in two-dimension plane in which robots are randomly distributed.The core of the first problem is to choose a reasonable corresponding target point for each robot.So a new method called artificial social position method which is put forward according to the minds of employers in society is proposed.Based on the self-organizing adjustment method,robots can change their next target points by the information they detected in real time and the three rules of artificial social position method,until they have their own points,and forming a valid,ordered and rational target formation.The second problem is motion planning problem(that is design of motion controller).Based on the characteristics of swarm-robot system model,this paper designs robots' artificial moment motion controller by artificial moment method.Its working principle is: to output robots' steering and to drive robots to reach their target points using artificial moment method after obtaining the initial position and the target position.Matlab simulation results indicate that the proposed method is feasible and effective.